OpenAI on Thursday launched a powerful new AI agent for its ChatGPT chatbot, enabling it to complete complex, multi-step tasks as the Microsoft-backed startup intensifies efforts to stay ahead in the rapidly evolving artificial intelligence race.

The new agentic tool marks a significant upgrade for ChatGPT, blending capabilities from OpenAI’s earlier features like “operator,” which interacts with websites, and “deep research,” which handles advanced queries requiring multiple steps.

Available immediately to users on ChatGPT’s Pro, Plus, and Team tiers, the AI agent can carry out a wide range of real-world tasks, such as selecting and ordering an outfit for a wedding based on dress code, weather, and user preferences. It does this through a virtual computer environment that can browse the web, manage data, and interact with third-party apps.

Users can also link applications such as Gmail and GitHub, allowing ChatGPT to pull relevant information from their accounts in response to prompts. This integration reflects a growing trend in AI development: building autonomous agents that can boost productivity and automate everyday tasks.

Tech giants including Microsoft, Salesforce, and Oracle are investing heavily in similar AI tools as they look to enhance business operations and reduce costs. The release of OpenAI’s agent signals a deepening of competition in this space, as companies seek to expand the role of AI from conversational assistants to autonomous digital workers.

The move comes amid surging interest in generative AI, with firms racing to deploy increasingly sophisticated tools across industries from customer service to software development. OpenAI’s latest release positions ChatGPT not just as a chatbot, but as a capable AI assistant with real-world utility.